This proposal seeks to elucidate the pathophysiology of and evaluate novel therapy for postischemic, acute renal failure (ARF) in humans. Delayed function (DF) of a cadaveric renal allograft (Tx) will serve as a prototypic model of ARF. We will examine 140 consecutive Tx recipients of whom half are predicted to manifest DF and severe hypofiltration. The remaining half, who will manifest prompt function (PF) and normofiltration will serve as controls. We wish to test four main hypotheses. Hypothesis #1 is that postischemic injury in those with DF lowers the GFR mainly by depressing the transcapillary hydraulic pressure gradient. GFR (insulin clearance) and its remaining four determinants will be evaluated 1-3 hr and 7 days after reperfusion of the Tx. Renal plasma flow (RPF) will be determined initially by Doppler flow meter and again on day 7 by a novel, non-invasive technique based on phase contrast, cine-MRI. Oncotic pressure will be determined by membrane osmometry. Glomeruli obtained by biopsy on each occasion will be subjected to a morphometric analysis and a hydrodynamic model of viscous flow to determine filtration surface area (S) abd hydraulic permeability (k). A model of ultrafiltration will be used. We seek to confirm that sustained ARF in humans is a consequence of dissipation of the net pressure for ultrafiltration. Hypothesis #2 is that depression in Tx recipients with DF vs PF is associated with activation of tubulo-glomerular feedback (TGF) and consequent afferent vasoconstriction. The fractional excretion of Li+ will be used as a surrogate for Na+ delivery to the macula densa. Impaired proximal Na+ reabsorption will be related to renovascular resistance and to polarity of proximal tubule cells in the serial biopsies. Cell polarity will be determined from the distribution of Na+/K+- ATPase and various cytoskeletal proteins of the plasma membrane using confocal microscopy. Hypothesis #3 is that enhanced paracellular flow through damaged tubules allows filtrate to leak back into the interstitium, thereby further lowering the clearance of dextrans of graded size to calculate the fraction of filtered insulin that leaks back. This will then be related to structural alterations in denudation of tubular basement membrane and expansion of their interstitium. Finally, we will conduct a controlled trial of insulin-like growth factor (IGF-1) vs placebo in all Tx recipients predicted to exhibit DF and ARF by a GFR<15 ml/min at the 1-3 hr study. rh-IGF-1 in restoring GFR by day 7 will be attributable to regeneration of an insulin-permeable proximal nephron lined by backleak and TGF-mediated afferent constriction.